Context
PhD Candidate: Amandine Dumont - amandine.dumont@uclouvain.be
Promoters: G. Gilquin & S. Granger
Period: 2013-2017
Research questions
Stenstrom & Svartvik (1994: 243) note that "native-like use of pauses, fillers and repeats is a quick way for foreign learners to improve their English language
proficiency", while Hasselgren (2002) shows that the use of smallwords like you know or I mean gives an impression of fluency. This underlines the importance, for learners of English, to use fluency markers, and to use them appropriately.
The aim of this thesis is to gain insight into the (dis)fluency behavior of learners of English, as compared to native speakers of English. The focus will be on
French-speaking learners of English and the approach will be integrated, in the sense of Gotz (2011).
- How can French-speaking learners' speech be characterized in terms of (dis)fluency and how does this compare to native speech?
- What fluency profiles can be identified among French-speaking learners and are these profiles valid for learners from other mother tongue
backgrounds? - What is the importance of idiolects in the measure of fluency across corpora of native and non-native speech?
Hypotheses
- Learners' fluency behavior differs from that of native speakers, both in terms of over-/underuse and misuse. It is expected that French-speaking
learners overuse (filled and silent) pauses, repeats and drawls, but tend tounderuse discourse markers (cf. Gilquin 2008). We also expect to find
(occasional or regular) misuse of certain (dis)fluency markers, e.g. interruption of closely-knit structures, inappropriate choice of marker,
unusual clustering of markers (cf. Gilquin & Granger 2011). - A number of fluency profiles can be established by examining the use of (dis)fluency markers in various corpora. Some of these profiles are shared
by learners from different mother tongue backgrounds, and even by native speakers. Others are limited to certain populations. - There is a great deal of variation in (dis)fluency among the different speakers represented in a corpus, hence the importance of considering
individual data in addition to aggregate data.
Main lines of investigation
The first one is descriptive and aims at giving a comprehensive description of the use of (dis)fluency markers in French-speaking learners English, and relating this to native behavior in comparable corpora of speech. A variety of markers will be investigated both from a quantitative and qualitative point of view.
On the basis of these descriptions, we will seek to identify fluency profiles among learners and native speakers. Each profile will correspond to a particular combination of features and will contribute to defining different speaker types. The profiles established for native speakers and French-speaking learners of English will then be compared to those established by Gotz (2011) for German learners, in an attempt to determine the degree of generalization of these profiles.
The comparison of the fluency profiles for French- and German-speaking learners will be the starting point of an analysis seeking to assess the role of transfer from the mother tongue in learners' use of (dis)fluency markers. Features that are specific to the French-speaking learners (and different from native speakers) will be considered as possibly due to L1 interference. In an attempt to confirm these hypotheses, the fluency behavior of native speakers of French will be considered (cf. WP 3) and compared with their behavior in English.
Finally, the issue of idiolects will be examined. While traditionally a corpus is considered as 'one big text', in which "the data obtained from (...) different speakers or writers are pooled" (Rietveld et al. 2004: 350), corpus linguists are increasingly aware of the fact that "corpora are inherently variable internally" (Gries 2006: 110). This is particularly true of learner corpora (Granger et al. 2009: 3). With a view to taking this variability into account, the data will also be approached from a more individual perspective, by looking at each text/speaker separately. We expect this approach to bridge the gap between learners and native speakers, with some learners performing better in terms of fluency than certain native speakers.
Publications
<to come>